当我在caffe上运行自己的图像时，它在迭代0，测试网(#0)处停止 [英] When I run my own images on caffe , it stop at Iteration 0, Testing net (#0)

问题描述

我跑了咖啡，得到了以下输出:

I ran caffe and got this output:

谁能告诉我问题出在哪里?
我将非常感激！

who can tell me what is the problem?
I will really appreciate!!

推荐答案

您的一个或多个标签值似乎无效，请参见

It seems like one (or more) of your label values are invalid, see this PR for information:

如果您的地面真理标签无效，则"SoftmaxWithLoss"将以静默方式访问无效内存[...]旧检查仅在DEBUG模式下有效，并且仅对CPU有效.

If you have an invalid ground truth label, "SoftmaxWithLoss" will silently access invalid memory [...] The old check only worked in DEBUG mode and also only worked for CPU.

请确保您的预测向量长度与您尝试预测的标签数量相符.

Make sure your prediction vector length matches the number of labels you try to predict.

从您的评论看来，您似乎在 "fc7" 仅预测1000个类别的概率.因此，"SoftmaxWithLoss"层尝试计算损耗以预测标签> 1000，并在概率阵列之外访问内存，从而导致分段错误.

From your comments, it seems like you have labels in the range 0..10575, but on the other hand, your classification layer, "fc7" only predicts probabilities for 1000 classes. Thus, "SoftmaxWithLoss" layer tries to compute the loss for predicting label l>1000, and access memory outside the probability array, resulting with a segmentation fault.

这篇关于当我在caffe上运行自己的图像时，它在迭代0，测试网(#0)处停止的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！